ARTICLE

DOI: 10.1038/s41467-018-04873-7 OPEN Revealing the immune perturbation of black phosphorus nanomaterials to macrophages by understanding the protein corona

Jianbin Mo 1,2, Qingyun Xie3, Wei Wei 1,2 & Jing Zhao 1

The increasing number of biological applications for black phosphorus (BP) nanomaterials has precipitated considerable concern about their interactions with physiological systems. 1234567890():,; Here we demonstrate the adsorption of plasma protein onto BP nanomaterials and the subsequent immune perturbation effect on macrophages. Using liquid chromatography tandem mass spectrometry, 75.8% of the proteins bound to BP quantum dots were immune relevant proteins, while that percentage for BP nanosheet–corona complexes is 69.9%. In particular, the protein corona dramatically reshapes BP nanomaterial–corona complexes, influenced cellular uptake, activated the NF-κB pathway and even increased cytokine secretion by 2–4-fold. BP nanomaterials induce immunotoxicity and immune per- turbation in macrophages in the presence of a plasma corona. These findings offer important insights into the development of safe and effective BP nanomaterial-based therapies.

1 State Key Laboratory of Coordination Chemistry, Institute of Chemistry and BioMedical Sciences, School of Chemistry and Chemical Engineering Nanjing University, Nanjing 210093, China. 2 State Key Laboratory of Pharmaceutical Biotechnology, School of Life Sciences, Nanjing University, Nanjing 210093, China. 3 Department of Orthopedics, Chengdu Military General Hospital, Chengdu 610083, China. These authors contributed equally: Jianbin Mo, Qingyun Xie. Correspondence and requests for materials should be addressed to W.W. (email: [email protected]) or to J.Z. (email: [email protected])

NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7

lack phosphorus (BP) crystals have existed for a long time, found to be free-standing with a lateral size of approximately 300 Bbut we have only recently begun to realize their true nm, and BPQDs were found to be approximately 5 nm in size potential from the perspective of two-dimensional (2D) (Fig. 2a, b). According to a statistical TEM analysis of 100 BPQDs nanomaterials1. BP thin film material has been produced from (Fig. 2b), the average lateral size of a BPQD is 5.7 ± 1.0 nm (values bulk BP crystals through the exfoliation method, and this material are expressed as the means ± SDs of triplicates). The BPQDs and exhibits unique qualities such as thickness-dependent band gap BPNSs were also characterized by X-ray photoelectron spectro- and high carrier mobility2,3. While many breakthroughs have scopy (XPS) and Raman spectroscopy. As shown in Fig. 2c, the been made regarding nanoelectronic applications for BP nano- similar prominent peaks from BPQDs and BPNSs can be 4 1 −1 materials , studies on extensive applications in biomedicine and attributed to one out-of-plane phonon mode (A g) at 356.8 cm biotechnology have become more common. Several groups have as well as two in-plane modes, B and A2, at 432.3 and 459.6 cm − 2g g reported that BP nanomaterials could serve as high-efficiency 1, respectively. Furthermore, similar survey XPS spectra for photosensitizers for photodynamic therapy and drug delivery BPQDs and BPNSs document the similar surface chemistry of – systems with ultra-high loading capacity5 8. Recently, potential these two BP nanomaterials (Fig. 2d). Taken together, the data inflammatory effects of BP nanomaterials have been reported9. indicate that size is the key difference between BPQDs and However, understanding of their actual biological effects on BPNSs. physiological systems is limited. Numerous studies have shown visual images of the corona on It has now been established that the surfaces of nanomaterials nanospheres or nanorods, but nanosheet coronas have rarely are immediately covered by a plasma protein corona when they been reported. Therefore, to provide additional qualitative come into contact with blood10,afinding that has led to the support for the stability of the isolated BP–corona complexes, redefinition of the biological properties of nanoparticles (NPs) we present TEM images of BPQDs and BPNSs with protein 11,12. Chen et al. elegantly demonstrated binding of the bovine corona (Fig. 2e, f). In BPNS–protein corona complexes, the serum albumin (BSA) protein corona to Au nanorods and surfaces of the BPNSs were strikingly different from their native revealed reduced damage to cell membranes13. Dawson et al. have forms, and the zeta-potential values of BPNSs decreased from made pivotal progress on understanding the role of the protein −18.1 to −8.4 mV (Fig. 2i), suggesting adsorption of plasma corona in the loss of targeting capabilities of NPs14. proteins onto the BPNSs. The size of BPNS–corona nanomater- The implications of nanomaterial–corona complexes are far ials was also investigated by dynamic light scattering (DLS) reaching, suggesting that the biological impacts of nanomaterials analysis (Fig. 2g and Supplementary Table 1). The results showed on a physiological system cannot simply be linked to the nature of that the average size of BPNSs changed from 338.4 ± 2.3 to 365.3 the nanomaterial alone. Therefore, for continued development of ± 5.9 nm. Unexpectedly, after the protein corona was formed on BP nanomaterials in biomedicine, studies on interactions between the surface of BPQDs, BPQDs were redefined from ultra-small nanomaterials and plasma proteins and the influence on immune nanosheets to spherical BPQD–corona complex NPs, and the NPs cells are urgent. Here we study the protein corona formed by appeared to be bulky particles with a diameter of approximately exposing different sizes of BP nanomaterials to plasma proteins 371.9 ± 9.1 nm (Fig. 2b, f). The zeta-potential values of BPQDs and demonstrated their immune perturbation effect on macro- decreased from −20.6 to −6.4 mV, and the “wet” diameter phage (Fig. 1). Our results also highlight the importance of increased from 5.6 ± 1.4 to 362.5 ± 5.6 nm (Fig. 2h, i and nanomaterial size on protein corona formation and could further Supplementary Table 1). Energy dispersive X-ray spectroscopic redefine BP nanomaterials and the biological response to BP (EDX) analysis further verified that BPQD was in the corona nanomaterials. complex (Supplementary Figure 1).

Results Identification of the plasma corona. According to the literature, Characterization of nanomaterials. First, BP nanosheets the formation of plasma proteins on the surfaces of nanomaterials (BPNSs) and BP quantum dots (BPQDs) were obtained according can occur almost instantly after nanomaterials enter the blood15. to reported methods5,7 and analysed by transmission electron The zeta-potential values for BP nanomaterials showed that the microscopy (TEM). From the TEM image results, BPNSs were high initial values (−20.6 mV for BPQDs and −18.1 mV for

Vessel

Cytokines BP nanomaterial

NF-κB

ROS

BP–corona Macrophage complex

Fig. 1 Summary diagram of immune perturbation of BP–corona complexes. Formation of BP nanomaterial–corona complexes in blood and their immunoregulation on macrophages

2 NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 ARTICLE

ace 2 BPNSs 1 A g 1000 A g BPQDs

800 B2g

600

400

Raman intensity 200

0

350400 450 500 550 Raman shift (cm–1)

bd f 25 N=100 O 1s BPNSs 20 N=100 20 5.7±1.0 nm 371.9±9.1 nm BPQDs 15 15 10 10

Number of particles 5 5 P 2s Number of particles 0 0 3456789 P 2p 100 200 300 400 500 600 700 Diameter (nm) Diameter (nm) Counts (s)

800 600 400 200 0 Binding energy (eV) BP gh i BP –corona 0 120 BPNSs 120 BPQDs BPNSs–corona complexes BPQDs–corona complexes 100 100 –5

80 80 –10 60 60 Number Number 40 40 –15

20 20 Zeta potential (mV) –20 BPQDs 0 0 BPNSs 120 240 360 480 600 4 5 6 7 300 360 420 Size (nm) Size (nm)

Fig. 2 Image of BP nanomaterials and corona complexes. a TEM image of a BPNS. Scale bar: 200 nm. b TEM image of BPQDs. Inset: statistical analysis of the size of 100 BPQDs determined by TEM. Scale bar: 200 nm. c Raman spectra of BPQDs and BPNSs. d XPS spectra of BPQDs and BPNSs. e TEM image of a BPNS–corona complex. Scale bar: 50 nm. f TEM image of BPQD–corona complex. Inset: statistical analysis of the size of 100 BPQD–corona complexes determined by TEM. Scale bar: 500 nm. g Dynamic light scattering (DLS) analysis of BPNSs and BPNS–corona complexes. h DLS analysis of BPQDs and BPQD–corona complexes. i Zeta potential of BP nanomaterials and BP–corona complexes. Values are expressed as the means ± SDs of triplicates

BPNSs) were reduced after entry (to −7.3 mV for BPQDs and the intensity of a fixed band, the 40 KD marker band in the first −13.7 mV for BPNSs in 0.5 h), suggesting that the proteins were lane) of several bands marked with an asterisk from the gels in adsorbed onto BP nanomaterials in a short time (Supplementary Fig. 3a, b. As shown in Fig. 3d, 55 and 60 KD proteins were main Figure 2). Then BP nanomaterials, including BPNSs and BPQDs, components in corona at low plasma concentration, while other were exposed to blood plasma of different concentrations for 4 h. proteins (i.e., 35, 70 and 180 KD protein) dominated the corona Plasma proteins in corona were then separated by sodium composition at high plasma concentration. These data suggested dodecyl sulphate-polyacrylamide gel electrophoresis (SDS- that the size of BP nanomaterials played a key role in determining PAGE). As Fig. 3a shows, with an increasing plasma concentra- the BP–corona complexes. tion, typical bands from BPNS–corona became more visible (i.e., Next, we used liquid chromatography tandem mass spectro- more same type proteins were enriched at higher concentrations). metry (LC-MS/MS) to qualitatively analyse components of the Notably, we found that high-abundant proteins on BPQDs were BP–protein corona (Supplementary Data 1, 2 and 3). Further changed when BPQDs interacted with plasma of different con- analysis with LC-MS/MS revealed that, while BPNSs were able to centrations. It has been suggested that the formation of corona adsorb 52 plasma proteins, BPQDs were capable of binding 96 was affected significantly by nanomaterial size.16,17. The different proteins. Of these proteins, 16 proteins appeared on both BPNSs protein-binding patterns from BPQD–corona in plasma of dif- and BPQDs (Fig. 3e). Subsequently, we analysed the difference in ferent concentrations revealed that competitive plasma proteins protein component among plasma proteins, BPQD–corona with an increasing adhesion at a higher concentration blood complex proteins and BPNS–corona complex proteins. First, this plasma could facilitate the desorption of proteins with lower analysis showed that the proteins on BPQDs and BPNSs do not binding affinity (Fig. 3b). Furthermore, we performed a more simply correspond to the relative protein concentrations in blood detailed study of the protein pattern to describe the observation plasma, as reported by other studies that focussed on other types associated with BPQD–corona. Thus, in Fig. 3c, d, we analysed of nanomaterials16,18. The bound proteins were further classified the relative densitometry (the intensity of each band is divided by by gene ontology (GO) analysis according to their biological

NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 process (Supplementary Figure 3). As shown in Fig. 3f, of the 186 negative charge of BP–corona complexes (Supplementary Fig- proteins in blood, 73 proteins in BPQD–corona complexes and 36 ure 2). Hence, an effective charge alone does not appear to be the proteins in BPNS–corona complexes were annotated as major driving force regulating the BP–protein interaction. In immune relevant proteins. These results indicated that 75.8% of addition, we discovered a significant enrichment of plasma the proteins bound to BPQDs (96 plasma proteins) were proteins with a high molecular weight (MW) on BPNSs rather immune relevant proteins, while that percentage for than on BPQDs, while proteins with a low MW were less BPNS–corona complexes (52 plasma proteins) is 69.9% (Fig. 3g). enriched on BPNSs than BPQDs (Supplementary Figure 5). These According to previous reports, negatively charged NPs primarily results indicated a distinct, protein size-dependent particle- attract positively charged proteins. By contrast, we observed that binding pattern. overall, of the proteins bound to the negatively charged BP nanomaterials, more than half of the proteins (64.93% for BPNSs In vitro cytotoxicity assay and cellular uptake efficiency and 62.00% for BPQDs) had a negative charge (pI < 7.4), experiments. Immune cells in the bloodstream (such as leuko- irrespective of their relative abundance in blood plasma cytes, dendritic cells and monocytes) and tissues (such as mac- (Supplementary Data 3 and Supplementary Figure 4). This result rophages) have a propensity to take up and clear certain was also confirmed by zeta-potential measurements, revealing a nanomaterials19. Once plasma proteins are adsorbed onto BP

ac2.0 170 5% plasma 130 * 10% plasma 1.6 20% plasma 100 40% plasma 70 80% plasma * 1.2 100% plasma 55 * 0.8 40

* Relative band intensity 0.4 35 25 0.0 5 10204080100

Control 55 KD 70 KD 37.4 KD 180 KD Plasma concentration (%)

b d 1.2 5% plasma 170 * 10% plasma 130 20% plasma 0.9 40% plasma 100 80% plasma 70 * 100% plasma 0.6 55 *

40 0.3 Relative band intensity 35 * 25 0.0 5 10204080100 35 KD 55 KD 60 KD 70 KD Control 180 KD Plasma concentration (%)

efgTotal proteins 200 250 Other Proteins on BPQDs Immune system process 186 ± 8 180 128 Proteins on BPNSs 200 39.2% 160 150

90 19.5% 80 16 36 73 ± 6 100 60 71.7%

Number of proteins Number of proteins 50 36 ± 5 75.8% 30 69.9% 0

Plasma Plasma

BPNS–coronacomplexes BPQD–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes

Fig. 3 Corona protein identification and classification. SDS-PAGE analysis of plasma protein obtained from a BPNS–corona complexes and b BPQD–corona complexes. The molecular weights of proteins in the standard ladder are reported on the left. Relative amounts of the most abundant proteins bound to c BPNSs and d BPQDs as calculated by the grey-scale value from the gels shown in a, b. Values are expressed as the means ± SDs of triplicates. e Statistical analysis of the total number of proteins identified by LC-MS/MS and f the number of proteins on BPQDs and BPNSs involved in immune system processes according to gene ontology (GO) analysis. Values are expressed as the means ± SDs of triplicates. g Analysis of the immune relevant protein fraction

4 NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 ARTICLE nanomaterials, some of these proteins could serve as opsonins to As previous literature has shown, the formation of a corona on influence cellular uptake and clearance by immune cells and thus nanomaterials is a dynamic process, and the components of a potentially affect distribution and delivery to the intended target human plasma corona are affected by foetal bovine serum sites. Since discovering the enrichment of immune proteins in (FBS)20,21. Thus it would be more suitable to carry out cell assays BP–corona complexes, we further focussed on the immunomo- in serum-free media to eliminate the interference of FBS. To dulatory effects of BP nanomaterial–protein corona complexes in verify the effect of serum concentration on cells, we first macrophages. In contrast to several studies that have reported performed an 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazo- nearly no cytotoxic effects of BP nanomaterials (including BPNSs lium bromide (MTT) assay to analyse the cell viability in different and BPQDs) in cell lines, our data showed that BPNSs and culture media (0 and 10% FBS) after a short treatment. As BPQDs produced slight cytotoxic effects on cell lines5,7, such as Supplementary Figure 8a shows, after 6 h, the cell viabilities human lung cancer cells (H1299), human normal hepatic cells (without any treatment) are 98% in serum-free media and 100% − (L0-2), human normal kidney cells (293T), human macrophage- in serum media. Even treated with nanomaterials (100 μgml 1) like cells (dTHP-1) and macrophages from peripheral blood (SC). in serum-free media, the cell viabilities are all >90%. We next Notably, when BPNSs and BPNSs were coated with the protein compared the cellular uptake of BP–corona complexes in corona, the cytotoxic effect was significantly reduced (Supple- different culture medium (0 and 10% FBS). The results of the mentary Figure 6 and 7). cellular uptake experiment showed that FBS had no significant To further analyse the clearance of BP nanomaterials by effect on cellular uptake efficiency during a short incubation macrophages, macrophage-like dTHP-1 cells were selected to period (Supplementary Figure 8b). Based on the above results, we evaluate the cellular uptake efficiency of BP–corona complexes. carried out further measurements using macrophage-like cells in

a 200 200 Control Control Plasma Plasma BPQDs BPNSs 160 BPQDs–corona complexes 160 BPNSs–corona complexes cells) cells) 4 120 4 120

80 80 Cellular uptake Cellular uptake (ng per 10 (ng per 10 40 40

136 1 3 6 Time (h) Time (h)

b DAPI Lyso Tracker Corona complexes Merged complexes BPQD–corona complexes BPNS–corona

c BPQDsBPQD–corona BPNSs BPNS–corona complexes complexes

Fig. 4 Cellular internalization of BP nanomaterials and corona complexes. a Cellular uptake of BP nanomaterials and BP–corona complexes. Macrophage- like dTHP-1 cells were treated with 150 μgml−1 nanomaterials for 1, 3 and 6 h. Cells treated with human plasma were set as the control. Values are expressed as the means ± SDs of triplicates. b, c Location of BP nanomaterials and BP–corona complexes in dTHP-1 cells. The cells were treated with 150 μgml−1 BP–corona complexes for 6 h and visualized using b fluorescence microscopy (scale bar: 10 μm) and c TEM (scale bar: 2 μm)

NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 serum-free media to eliminate the interference of FBS. To clarify 104 cells after 6 h) than that of cells treated with BPNSs (130.8 ng the association between the corona and cellular uptake, we next per 104 cells after 6 h), which was the reverse of the uptake investigated the uptake efficiency of BPNSs, BPQDs and pattern for graphene oxide nanosheets22. Typically, higher uptake BP–corona complexes in dTHP-1 cells using inductively coupled levels of BP–corona complexes (175.0 ng per 104 cells for plasma mass spectrometry (ICP-MS). As shown in Fig. 4a, the P BPNS–corona complexes and 145.1 ng per 104 cells for content of cells incubated with BPQDs was higher (137.8 ng per BPQD–corona complexes) were detected after 6 h, indicating

a 100 100 100

80 * 80 80

) * * * ) ) –1 –1

–1 * 60 60 60 * (pg ml

β 40 40 40 IL-8 (pg ml IL-6 (pg ml IL-1 20 20 20

0 0 0 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 Nanomaterials (μg ml–1) Nanomaterials (μg ml–1) Nanomaterials (μg ml–1)

BPQDs BPNSs BPQDs BPNSs BPQDs BPNSs

BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes

75 75 125 ** * 60 60 100 ) ) ) –1 –1 –1 45 45 75 * * (pg ml

* γ 30 * 30 50 IL-9 (pg ml IL-10 (pg ml IFN- 15 15 25

0 0 0 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 Nanomaterials (μg ml–1) Nanomaterials (μg ml–1) Nanomaterials (μg ml–1)

BPQDs BPNSs BPQDs BPNSs BPQDs BPNSs

BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes

b 75 75 75

60 60 60

) ** * ) ) –1 * –1 * * –1 * 45 45 45 (pg ml

β 30 30 30 IL-6 (pg ml IL-8 (pg ml IL-1 15 15 15

0 0 0 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 Nanomaterials (μg ml–1) Nanomaterials (μg ml–1) Nanomaterials (μg ml–1)

BPQDs BPNSs BPQDs BPNSs BPQDs BPNSs

BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes 75 125 125

60 100 * 100 ) ) ) * –1 –1 –1 45 * * 75 75 (pg ml 30 50 γ 50 IL-9 (pg ml IL-10 (pg ml IFN- 15 25 25

0 0 0 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 Nanomaterials (μg ml–1) Nanomaterials (μg ml–1) Nanomaterials (μg ml–1)

BPQDs BPNSs BPQDs BPNSs BPQDs BPNSs

BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes BPQD–coronacomplexes BPNS–coronacomplexes

Fig. 5 Cytokine secretion of different macrophages. Macrophage-like dTHP−1 cells (a) or SC cells (b) were treated with 50 μgml−1 BP nanomaterials and an increasing concentration of corona complexes (12.5, 25 and 50 μgml−1) for 6 h. Values are expressed as the means ± SDs of triplicates. Statistical significance is assessed by Student’s t test. *p < 0.05, **p < 0.01

6 NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 ARTICLE the facilitation of BP nanomaterial internalization by the TEM were combined to study the intracellular localization of corona. BP–corona complexes. As shown in Fig. 4b, c, BP–corona com- plexes were internalized into macrophages by endocytosis, and the primary final location of the corona complexes was the Cellular locations. To further investigate the intracellular beha- fl viour, we added human serum albumin conjugated to fluorescein cytoplasm. Moreover, the green uorescence intensity of the fl BPNS–corona complexes was stronger than that of the isothiocyanate (HSA-FITC) to plasma to create uorescent cor- – fi ona complexes (Supplementary Figure 9). First, we studied the BPQD corona complexes, suggesting that BPNSs modi ed with stability of fluorescent corona complexes in macrophage-like cell plasma protein exhibited a superior ability to accumulate in macrophage-like cells compared with BPQDs. Nanomaterial size lysates to simulate the macrophage cell environment. As shown in fl fi Supplementary Figure 10, only 4.3% of HSA-FITC was dis- in uences the ef ciency of subsequent cellular uptake, probably sociated from BPNS–corona complexes and 6.3% of HSA-FITC by controlling the formation of the protein corona coating. was dissociated from BPQD–corona complexes after 10 h, indi- cating the stability of fluorescent corona complexes in intracel- Proinflammatory effect of BP–corona complexes in macro- lular environments. In this way, we were able to label BP–corona phages. Presently, increasing studies have shown that protein complexes with fluorescence tags and determine their cellular corona formation was an important pathway for nanomaterials to location in dTHP-1 cells. Thus, fluorescence microscopy and stimulate the inflammatory response23,24. Other nanomaterial

a 100 75 * * 80 * * ) 60 –1 ) 60 –1 45 (pg ml α 40 30 NO (pg ml 20 15

0 0 0 50 12.5 25 50 50 12.5 25 50 0 50 12.5 25 50 50 12.5 25 50 Nanomaterials (μg ml–1) Nanomaterials (μg ml–1)

BPQDs BPNSs BPQDs BPNSs

complexes complexes BPQD–corona BPNS–corona BPQD–coronacomplexes BPNS–coronacomplexes

b 150 150 Nanomaterials (0 μg/ml)+LPS Nanomaterials (0 μg/ml)+LPS Nanomaterials (12.5 μg/ml)+LPS Nanomaterials (12.5 μg/ml)+LPS 120 Nanomaterials (25.0 μg/ml)+LPS 120 Nanomaterials (25.0 μg/ml)+LPS ) TNF- Nanomaterials (50.0 μg/ml)+LPS ) Nanomaterials (50.0 μg/ml)+LPS –1 –1 90 90 (pg ml (pg ml α 60 α 60 TNF- 30 TNF- 30

0 0 Control BPQD–corona BPQDs Control BPNS–corona BPNSs complexes complexes

c 300 300 BPQDs–corona complexes BPNSs–corona complexes BPQDs BPNSs 250 LPS 250 LPS Control Control

200 200

150 150 ROS production (%) ROS production (%) 100 100

0 30 60 90 120 150 0 30 60 90 120 150 Time (min) Time (min)

Fig. 6 Immune perturbation by BP–corona complexes in macrophages. a NO and TNF-α generation in dTHP-1 cells. Cells were treated with 50 μgml−1 BP nanomaterials and an increasing concentration of corona complexes (12.5, 25 and 50 μgml−1) for 6 h. Values are expressed as the means ± SDs of triplicates. Statistical significance is assessed by Student’s t test. *p < 0.05, **p < 0.01. b TNF-α secretion in dTHP-1 cells. LPS (20 ng ml−1) was added to the macrophages, which had been pretreated with BP nanomaterials or BP–corona complexes for 6 h. Values are expressed as the means ± SDs of triplicates. c ROS overproduction in macrophages exposed to 50 μgml−1 BP nanomaterials or BP−corona complexes for 6 h. LPS (20 ng ml−1) was set as the positive control. Values are expressed as the means ± SDs of triplicates

NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications 7 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 systems, including polymer NPs, gold NPs, superparamagnetic of TNF-α in dTHP-1 cells treated with LPS was significantly iron oxide NPs, etc., have been reported to induce proin- inhibited. For example, after dTHP-1 cells were pretreated with – − flammatory cytokines in immune cells21,25 28. To test whether 50 μgml 1 BPQD–corona complexes and then incubated with 20 − the protein corona bound to BP nanomaterials could increase the ng ml 1 LPS, the secretion of TNF-α was decreased to 49.83 from − secretion of cytokines, macrophage-like cells were treated with 95.16 pg ml 1 (without pretreatment with BPQD–corona com- either BP nanomaterials or BP–corona complexes. Subsequently, plexes). These results supported the immunotoxicity of the culture supernatants were collected, and the amount of BP–corona complexes to macrophages in vitro. inflammatory cytokines was detected by enzyme-linked immu- ROS are an important part of the primary immune defence in nosorbent assay (ELISA). Significant changes in cytokine secre- macrophages against foreign materials. In addition, the expres- tion were observed in macrophage cells (dTHP-1 cells and SC sion of ROS has also been recognized as a major toxicological cells) treated with corona complexes. As shown in Fig. 5a, paradigm of nanomaterials40,41. The ROS level induced by BP–corona complexes induced an obvious increase in four BP–corona complexes was tested by dichlorofluorescein (DCF) inflammatory cytokines: interleukin (IL)−1β, IL-6, IL-8, and fluorescence assay (Fig. 6c). ROS production was significantly interferon (IFN)-γ. In particular, these cytokine secretions in increased in a time-dependent manner in macrophage-like cells dTHP-1 cells were increased 2–4-fold above background level. treated with BP nanomaterials or BP–corona complexes. However, compared with BP–corona complexes, BP nanoma- Compared with BP nanomaterials, corona complexes induced terials had no significant effect on the secretion of cytokines. IL-9 much higher ROS production. For example, BPQD– and and IL-10 expression was also not obviously changed upon BPNS–corona complexes triggered ROS overproduction in exposure to BP nanomaterials or BP–corona complexes. More- dTHP-1 cells at approximately 185.8% and 215.2% at 150 min, over, a similar trend in cytokine expression was observed in which was much higher than that elicited by BPQDs (143.7%) macrophages from peripheral blood (SC cells). The secretion of and BPQNs (159.9%). These findings demonstrated that BP IL-1β, IL-6, IL-8, IL-9 and IL-10 in SC cells was also enhanced by nanomaterials exhibited enhanced immune perturbations in BP–corona complexes (Fig. 5b). However, IFN-γ expression was macrophages because of the formation of plasma coronas. not significantly changed upon exposure to BP nanomaterials or corona complexes. Taken together, the higher secretion of Discussion inflammatory cytokines observed in different human macrophage So far, different roles of nanomaterial–protein corona have been cells are in support of the proinflammatory effect of BP nano- defined, including reduction of toxicity to cells, loss of targeting materials in the presence of a plasma corona. capabilities, extension of blood circulation time, alterations in Generally, the nuclear factor (NF)-κB pathway has been organ distribution and immunostimulatory and immunosup- implicated as a key factor in inflammatory effects induced by – pressive effects42 47. As BP-based nanomedicine is of increasing foreign materials29. The activation of NF-κB results in lκB importance, immunotoxicological assessments of BP nanoma- degradation and nuclear translocation of the p65 dimer30. After terials are urgently required. Hence, we carried out this study to treatment with BP nanomaterials or corona complexes, the provide a detailed investigation of the plasma protein corona cytoplasmic and nuclear extracts of dTHP-1 cells were analysed around BP nanomaterials and its impact on macrophages. Our by western blot (WB). The results in Supplementary Figure 11 results show that the size of 2D BP nanomaterials can influence show lκB degradation and nuclear translocation of p65. These the identity and quantity of plasma proteins that form the corona, changes elicited by BP nanomaterials were much lower than those and in return, the corona could define the shape of ultra-small 2D observed with corona complexes. To further understand the nanomaterial–corona complexes but does not change the shape of relationship between immune relevant proteins on BP nanoma- large 2D nanomaterials. Owing to the enrichment of immune terials and immune effects, we carefully searched the reported proteins and other opsonins on the surface of BP nanomaterials, literatures for the effect of every immune protein identified by the uptake efficiency of BP–corona complexes by macrophages is GO analysis and speculated that BP–corona complexes have more higher, and the size-dependent cellular uptake pattern of potential immune responses (Supplementary Table 2, 3). BP–corona complexes is different from that of native nanoma- terials. Our data also highlight the importance of the protein corona to the immune perturbation effect on macrophages. As Immune perturbation of BP–corona complexes in macro- cellular uptake of BP–corona complexes is enhanced, BP nano- phages. According to previous reports, immunotoxicity of materials bound to opsonins have a distinct impact on the nanomaterials means the effect of nanomaterials on immune cell proinflammatory and immune perturbation effect in macro- functions, including inflammatory cytokines secretions, reactive phages. This reveals the pivotal role of the corona in potential oxygen species (ROS) overexpression, phagocytic activity, pro- – innate immune and inflammatory diseases. To further develop BP liferative activity, etc31 33. Evaluation in vitro of the immuno- nanomaterials for effective biological applications, more studies toxicity of nanomaterials has been reported34,35. In addition, the must aim to modify BP nanomaterials to control the formation of induction of nitric oxide (NO) and tumour necrosis factor (TNF)- the protein corona and its subsequent influence on compatibility α production are molecular markers for the immunotoxicity of for better therapy. nanomaterials36,37. To elucidate the immune perturbation in macrophages, we first studied the induction effect of BP–corona complexes on NO and TNF-α. As shown in Fig. 6a, TNF-α and Methods NO production were provoked by BP nanomaterials and corona Material and cell lines. BP crystals were purchased from HWRK Chemical Co., complexes. The increase in TNF-α and NO expression induced by Ltd. (Beijing, China) and stored in a N2 glovebox. N-methyl pyrrolidone (NMP) was purchased from Aladdin Chemistry Co., Ltd (Shanghai, China). HSA-FITC corona complexes was two times greater than that of BP nano- was purchased from Biosynthesis Biotechnology Co., Ltd (Beijing, China). Human materials. The bacterial endotoxin lipopolysaccharide (LPS) is lung cancer cells (H1299), human normal hepatic cells (L0-2), human embryonic known to induce the release of inflammatory factors in macro- kidney cells (293T), human monocytic leukaemia cells (THP-1) and human phages25,38,39. To analyse the effect of corona complexes on macrophages from peripheral blood (SC) were purchased from American Type Culture Collection (Manassas, VA). All cell lines were not listed by International macrophage function, we further evaluated stress in macrophage- Cell Line Authentication Committee as cross-contaminated or misidentified (v8.0, like cells caused by foreign materials. As the results of Fig. 6b 2016). All of the cell lines were authenticated by STR typing and confirmed to be show, after pretreatment with BP–corona complexes, the release mycoplasma-free by KeyGEN BioTECH Co., Ltd (Nanjing, China). All cells were

8 NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 ARTICLE cultured in 1640 or Dulbecco’s modified Eagle’s medium (DMEM) media (Gibco) _02) in FASTA format. In this study, the top 10 blast hits with E-values <1e-3 for − − with 10% FBS (Gibco), 100 units ml 1 penicillin, and 50 units ml 1 streptomycin at each query protein were retrieved and loaded into Blast2GO (Version 3.3.5) for GO 53 37 °C in a CO2 incubator (95% relative humidity, 5% CO2). Macrophage-like THP- mapping and annotation. The GO analysis was carried out with Blast2GO . All 1 (dTHP-1) cells were differentiated from THP-1 cells by treatment with 100 μgml experiments were conducted at least twice to ensure the reproducibility of the − 1 phorbol myristate acetate for 48 h. All reagents were purchased from Sigma results. unless otherwise indicated. Cytotoxicity assay. The cytotoxicities of BP nanomaterials and BP–corona Human plasma. Blood was taken from ten different apparently healthy donors complexes to cells were determined by MTT assay. Cells (2 × 104 cells per well) (according to the Declaration of Helsinki) who had provided informed consent for were seeded in 96-well tissue culture plates for 24 h and then incubated with 100 μl blood collection and subsequent analysis and stored in tubes containing EDTA to of fresh medium (0% FBS) containing different concentrations of nanomaterials − prevent blood clotting. The blood samples were labelled anonymously and could (100, 50, 25, 12.5 μgml 1) for 48 h. Subsequently, 30 μl per well of MTT solution not be traced back to a specific donor. The samples were centrifuged for 5 min at (KeyGEN BioTECH Co., Ltd, Nanjing, China) was added and incubated for 4 h. 1700 × g to pellet red and white blood cells. The plasma supernatant was pooled After incubation, the culture medium was removed and replaced with 150 μl per and stored at −80 °C. This plasma was used exclusively throughout the study. well dimethyl sulphoxide. The colour intensity of the medium was measure at 550 nm using a microplate reader (Tecan Infinite M1000 PRO) to calculate the cell viability. In addition, we also analysed the cytotoxicities of BP nanomaterials and Preparation of BPNSs and BPQDs. BPNSs and BPQDs were synthesized BP–corona complexes in 1640 or DMEM media (10% FBS) following the methods according to the previous reports with some slight changes5,7. Briefly, 25 mg of the above. All experiments were conducted at least twice to ensure the reproducibility BP power was added to 25 ml of NMP and sonicated in ice bath for 10 h at 600 W. of the results. Half of the resulting brown suspension was centrifuged at 1700 × g for 8 min to remove the residual unexfoliated particles, and the supernatant containing BPNSs was collected for further use. The other half of resulting brown suspension was Cellular uptake of BP nanomaterials and corona complexes. Macrophage-like − sonicated in an ultrasonic bath continuously for 15 h at 150 W. The dispersed dTHP-1 cells (1.2 × 106 cells ml 1) were seeded in 6-well tissue culture plates and mixture was kept at 0 °C. The resulting sample solution was centrifuged for 20 min grown to ~90% confluence. BPQDs and BPNSs (200 μg) were incubated with 200 μl at 5200 × g, and the supernatant containing BPQDs was gently decanted. Next, the of 100% plasma for 4 h at 4 °C. Corona complexes were obtained as described BPQD products could be collected by centrifugation, and the precipitate was above and resuspended in 1 ml 1640 media (0% FBS) to prepare for cell culture. repeatedly rinsed with water and resuspended in an aqueous solution. Meanwhile, 200 μg of free BPQDs and BPNSs were resuspended in 1 ml of DMEM (0% FBS). Growth media in plates was replaced with media containing nanoma- terials and corona complexes. Cells were incubated with nanomaterials for 1, 3 or 6 Protein–nanomaterial interaction experiments. BPQDs and BPNSs were incu- h at 37 °C. Finally, the medium was removed from the wells, and the cells were bated with different human plasma concentrations (5, 10, 20, 40, 80 and 100%) at washed three times with PBS to clear away nanomaterials outside of the cells. Next, 4 °C for 4 h, and plasma solutions were diluted with phosphate-buffered saline 200 μl of 1% Triton X-100 in 0.1 M NaOH solution was added to lyse the cells. (PBS) keeping the final nanomaterial concentration constant. To obtain protein After treatment, P content in the lysates was determined using ICP-MS. In addi- corona complexes, the mixed solutions were centrifuged to pellet particle–protein tion, we also measured the cellular uptake efficiency of BP–corona complexes in complexes. The precipitate was then resuspended three times with 500 μl PBS and 1640 media (10% FBS) following the methods above. All experiments were con- centrifuged again for 20 min at 15,000 × g at 4 °C. ducted at least twice to ensure the reproducibility of the results.

Characterization of nanomaterials. BPNSs, BPQDs and BP nanomaterial–protein Preparation of fluorescent BP–corona complexes and stability analysis. corona complex solutions were dispersed onto holey carbon film on copper grids BPQDs and BPNSs (200 μg) were incubated with 200 μl of 100% plasma (mixed and then observed by TEM on a JEOL JEM-1200EX at an acceleration voltage of with 10 μg HSA-FITC) for 4 h at 4 °C. To obtain protein corona complexes, the 120 kV. EDX was conducted on the TEM. mixed solution was centrifuged to pellet the corona complexes. The precipitate was then resuspended three times in 500 μl PBS and centrifuged again for 20 min at Zeta potential and DLS measurement. BP nanomaterials (BPQDs and BPNSs) 15,000 × g at 4 °C. The green fluorescence intensity of the BPNS–protein corona were incubated with plasma at 4 °C for 4 h. The products were washed three times complexes was tested by SDS-PAGE. with 500 μl PBS to thoroughly remove unbound plasma protein and centrifuged The stability of fluorescent BP–corona complexes was determined according to again for 20 min at 15,000 × g at 4 °C to collect BP–corona complexes. After previous literature36. Briefly, fluorescent BP–corona complexes were suspended in resuspension in PBS, the zeta potential and size of BP–corona complexes were the dTHP-1 cells lysate with shaking at 37 °C. At specific times during incubation, measured on a Nano-ZS instrument (Malvern Instruments Limited) and DLS the fluorescence intensity of the supernatant was measured by microplate reader (Brookhaven BI-200SM). In addition, the zeta potential and size of BPQDs and (Tecan Infinite M1000 PRO) with the excitation and emission wavelengths set to BPNSs were also measured in PBS. All experiments were conducted at least twice to 490 and 520 nm, respectively, after centrifugation to remove the corona complexes. ensure the reproducibility of the results. The fluorescence intensity of the corona complexes was also detected as the control. All experiments were conducted at least twice to ensure the reproducibility of the results. SDS-polyacrylamide gel electrophoresis. SDS-PAGE was carried out according to standard procedures48. Corona complexes were separated by 12% SDS-PAGE using the SDS-PAGE Preparation Kit (Sangon Biotech Co., Ltd, Shanghai, China). Intracellular localization of corona complexes. For fluorescence microscopy, Then proteins were visualized by staining with Coomassie brillian blue R-250. All dTHP-1 cells were first seeded onto 2-cm culture dishes at a density of 8.0 × 104 − experiments were conducted at least twice to ensure the reproducibility of the cells ml 1 for 24 h. Fluorescent BP–corona complexes (200 μg) were obtained results. following the methods above and then resuspended in 1 ml DMEM (0% FBS) to prepare for cell culture. Cells were incubated with nanomaterials for 6 h at 37 °C, and then the medium was removed from the dish. After rinsing with PBS three Protein identification and classification. The BP–corona complexes were col- times to remove residual nanomaterials, cells were immobilized with 1 ml 4% lected by the above methods. Then one volume of SDT buffer (4% SDS, 100 mM paraformaldehyde for 15 min. Then the cell nucleus and lysosome were stained dithiothreitol (DTT), 150 mM Tris-HCl pH 8.0) was added, and the solution was with 4,6-diamidino-2-phenylindole and Lyso Tracker, respectively (Beyotime boiled for 15 min and centrifuged at 14,000 × g for 20 min. Digestion of protein Biotechnology Co., Ltd, Shanghai, China); the cells were monitored using a (200 μg for each sample) was performed according to the FASP procedure fluorescence microscope (ZEISS). described by Mann et al49. In all, 200 μl UA buffer (8 M Urea, 150 mM Tris-HCl, For TEM, dTHP-1 cells were cultivated in 6-well plates at 1.0 × 106 cells per well pH 8.0) were added to removed the detergent, DTT and other low-molecular for 24 h. BPQDs and BPNSs (200 μg) were incubated with 200 μl of 100% plasma components by ultrafiltration (Microcon units, 30 KD). Reduced cysteine residues for 4 h at 4 °C. Corona complexes were obtained as described above and were blocked with 100 μl iodoacetamide (0.05 M in UA buffer) and the samples resuspended in 1 ml 1640 media (0% FBS) to prepare for cell culture. Meanwhile, were incubated for 20 min in the dark. The filter was washed with 100 μl UA buffer 200 μg of free BPQDs and BPNSs were resuspended in 1 ml DMEM (0% FBS). for three times and then with 100 μlNHHCO (25 mM) twice. Trypsin (3 μg, 4 3 Cells were incubated with nanomaterials for 6 h at 37 °C, and the medium was Promega) in 40 μlNHHCO (25 mM) was added to digest the proteins at 37 °C for 4 3 removed from the dish. After rinsing with PBS three times to remove the residual 12 h, and the resulting peptides suspension was collected by filtration. The content nanomaterials, cells were transferred to fixative, stained, sliced and imaged by TEM of peptide was determined with ultraviolet light spectral density at 280 nm using an (Hitachi HT7700) according to previous literature19. All experiments were extinction coefficient of 1.1 of 0.1% solution. Each fraction was injected into a Q- conducted at least twice to ensure the reproducibility of the results. Exactive mass spectrometer (Thermo Scientific) for LC-MS/MS analysis. The MS/ MS spectra were searched with MASCOT engine (Matrix Science, London, UK; version 2.4). The following option was used: peptide mass tolerance = 20 ppm, MS/ Cytokine secretion assay. To analyse the cytokine secretion by different MS tolerance = 0.1 Da, enzyme = Trypsin, missed cleavage = 2, fixed modification: macrophage-like cells, the cells were seeded onto 96-well plates at 1.0 × 105 cells per – carbamidomethyl (C), and variable modification: oxidation (M)50 52. The identi- well for 24 h. BPNSs, BPQDs and corona complexes were resuspended in culture fied proteins were retrieved from the UniProtKB human database (Release 2017 media (0% FBS), added to each well at different concentrations (0, 12.5, 25, 50 μg

NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications 9 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7

− ml 1) and incubated for 6 h. Then the supernatant was collected and the cytokines 7. Sun, Z. et al. Ultrasmall black phosphorus quantum dots: synthesis and use as (IL-1β, IL-6, IL-8, IL-9, IL-10, IFN-γ and TNF-α) and NO were measured using photothermal agents. Angew. Chem. Int. Ed. 127, 11688–11692 (2015). ELISA kits (R&D Systems, CA, USA). 8. Shao, J. et al. Biodegradable black phosphorus-based nanospheres for in vivo To analyse the immune perturbation effect of corona complexes and BP photothermal cancer therapy. Nat. Commun. 7, 12967 (2016). nanomaterials, dTHP-1 cells were seeded onto 96-well plates at 1.0 × 105 cells per 9. Qu, G. et al. Chemical modification improves biocompatibility of black well for 24 h. Nanomaterials were resuspended in serum-free media, added into phosphorus nanosheets. Angew. Chem. Int. Ed. 56, 14488 (2017). −1 each well at different concentrations (0, 12.5, 25, 50 μgml ) and incubated for 6 h. 10. Salvati, A. Biomolecular coronas provide the biological identity of nanosized Then the supernatant was removed, and the cells were washed with PBS three materials. Nat. Nanotechnol. 7, 779 (2012). −1 times. Subsequently, macrophages were further treated with LPS (20 ng ml ) for 6 11. Lynch, I. & Dawson, K. A. Protein-nanoparticle interactions. Nano Today 3, h. Finally, the supernatants were collected, and the cytokines were determined with 40–47 (2008). ELISA kits. All experiments were conducted at least twice to ensure the 12. Monopoli, M. P., Åberg, C., Salvati, A. & Dawson, K. A. Biomolecular coronas reproducibility of the results. provide the biological identity of nanosized materials. Nat. Nanotechnol. 7, 779–786 (2012). Measurement of ROS in macrophages. The effect of BP nanomaterials and 13. Wang, L. et al. Revealing the binding structure of the protein corona on gold corona complexes on intracellular ROS generation in macrophage cells was nanorods using synchrotron radiation-based techniques: understanding the detected with a Reactive Oxygen Species Assay Kit (KeyGEN BioTECH Co., Ltd, reduced damage in cell membranes. J. Am. Chem. Soc. 135, 17359–17368 Nanjing, China) following the manufacturer’s protocol. Briefly, dTHP-1 cells were (2013). seeded onto 96-well plates at 1.0 × 106 cells per well for 24 h, and then the culture 14. Salvati, A. et al. Transferrin-functionalized nanoparticles lose their targeting medium was replaced with 100 μl of DCF-DA/DMEM44 medium (final DCF capabilities when a biomolecule corona adsorbs on the surface. Nat. concentration is 10 μM) and incubated at 37 °C for 20 min. Then BP nanomaterials – − − Nanotechnol. 8, 137 143 (2013). (100 μgml 1) and corona complexes (100 μgml 1) were added to the wells. The 15. Hadjidemetriou, M. & Kostarelos, K. Nanomedicine: evolution of the fluorescence intensity of DCF was measured by microplate reader (Tecan Infinite nanoparticle corona. Nat. Nanotechnol. 12, 288–290 (2017). M1000 PRO) with the excitation and emission wavelengths at 488 and 525 nm, 16. Sund, J., Alenius, H., Vippola, M., Savolainen, K. & Puustinen, A. Proteomic fl respectively, and the uorescence intensity was used to evaluate the ROS level in characterization of engineered nanomaterial-protein interactions in relation to macrophage cells. All experiments were conducted at least twice to ensure the surface reactivity. ACS Nano 5, 4300 (2011). reproducibility of the results. 17. Gagner, J. E., Qian, X., Lopez, M. M., Dordick, J. S. & Siegel, R. W. Effect of gold nanoparticle structure on the conformation and function of adsorbed Western blot analysis. Macrophage-like dTHP-1 cells were treated with BP proteins. Biomaterials 33, 8503–8516 (2012). − nanomaterials (100 μgml 1) and BP–corona complexes for 6 h. Then the cells were 18. Tenzer, S. et al. Nanoparticle size is a critical physicochemical determinant of washed with PBS three times and collected. The cytoplasm and nuclear extracts the human blood plasma corona: a comprehensive quantitative proteomic were prepared using the Minute™ Cytoplasmic and Nuclear Fractionation Kit analysis. ACS Nano 5, 7155–7167 (2011). (Invent Biotechnologies, Inc), and the protein concentration was determined by the 19. Walkey, C. D., Olsen, J. B., Guo, H., Emili, A. & Chan, W. C. Nanoparticle size Super-Bradford Protein Assay Kit (CWBiotech, Inc, Beijing, China). The expres- and surface chemistry determine serum protein adsorption and macrophage sion levels of lκB in cytoplasm extracts and p65 in nuclear extracts were determined uptake. J. Am. Chem. Soc. 134, 2139 (2012). by WB according to previous literature54. Briefly, the extracts were first separated 20. Walczyk, D., Bombelli, F. B., Monopoli, M. P., Lynch, I. & Dawson, K. A. What by 12% SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Bio- the cell “sees” in bionanoscience. J. Am. Chem. Soc. 132, 5761–5768 (2010). Rad, CA, USA). The membrane was blocked with 5% BSA in PBS at 25 °C for 2 h 21. Yan, Y. et al. Differential roles of the protein corona in the cellular uptake of and then incubated with antibodies at 4 °C overnight. The expression of β-actin nanoporous polymer particles by monocyte and macrophage cell lines. ACS was used as the internal standard. The uncropped WB images were included in Nano 7, 10960 (2013). Supplementary Figure 12. The sources of primary antibody against the following 22. Mu, Q. et al. Size-dependent cell uptake of protein-coated graphene oxide β protein were: -actin (A1978, 1:10,000 for WB) from sigma; p65 in nuclear nanosheets. ACS Appl. Mater. Interfaces 4, 2259–2266 (2012). κ (ab16502, 1:1000 for WB), and l B (15095, 1:1000 for WB) from Abcam. The 23. Elsabahy, M. & Wooley, K. L. Cytokines as biomarkers of nanoparticle appropriate secondary antibodies (1:10,000 for WB) were purchased from Abcam. immunotoxicity. Chem. Soc. Rev. 42, 5552 (2013). 24. Neagu, M. et al. Protein bio-corona: critical issue in immune nanotoxicology. Statistical analysis. All the data are expressed as mean ± standard deviation. Arch. Toxicol. 91, 1031 (2017). Differences between different experimental groups were analysed by two-tailed 25. Deng, Z. J., Liang, M., Monteiro, M., Toth, I. & Minchin, R. F. Nanoparticle- Student’s t test. One-way analysis of variance was used in multiple group com- induced unfolding of fibrinogen promotes Mac-1 receptor activation and parisons. Statistical analysis was performed using SPSS statistical program version inflammation. Nat. Nanotechnol. 6,39–44 (2011). 13 (SPSS Inc., Chicago, IL). Differences with p < 0.05 (*) or p < 0.01 (**) were 26. Chen, F. et al. Complement proteins bind to nanoparticle protein corona and considered statistically significant. undergo dynamic exchange in vivo. Nat. Nanotechnol. 12, 387 (2017). 27. Escamilla-Rivera, V. et al. Protein corona acts as a protective shield against fl Data availability. All data are available from the corresponding authors upon Fe3O4-PEG in ammation and ROS-induced toxicity in human macrophages. – reasonable request. The LC-MS/MS data have been deposited to the Proteo- Toxicol. Lett. 240, 172 184 (2016). meXchange Consortium via the PRIDE55 partner repository with the data set 28. Wan, S. et al. The “sweet” side of the protein corona: effects of glycosylation identifier PXD009645. on nanoparticle-cell interactions. ACS Nano 9, 2157–2166 (2015). 29. Li, Q. & Verma, I. M. NF-κB regulation in the immune system. Nat. Rev. Immunol. 2, 975–975 (2002). Received: 29 November 2017 Accepted: 30 May 2018 30. Hayden, M. S., West, A. P. & Ghosh, S. NF-κB and the immune response. Oncogene 25, 6758 (2006). 31. Tulinska, J. et al. Immunotoxicity and genotoxicity testing of PLGA-PEO nanoparticles in human blood cell model. Nanotoxicology 9,33–43 (2013). 32. Shahbazi, M. A. et al. The mechanisms of surface chemistry effects of mesoporous silicon nanoparticles on immunotoxicity and biocompatibility. References Biomaterials 34, 7776–7789 (2013). 1. Li, L. et al. Black phosphorus field-effect transistors. Nat. Nanotechnol. 9, 372 33. Luo, Z. et al. Surface functionalized mesoporous silica nanoparticles with (2014). natural proteins for reduced immunotoxicity. J. Biomed. Mater. Res. A 102, 2. Zhou, Q., Chen, Q., Tong, Y. & Wang, J. Light-induced ambient degradation 3781–3794 (2015). of few-layer black phosphorus: mechanism and protection. Angew. Chem. Int. 34. Dan, P. Immunotoxicity derived from manipulating leukocytes with lipid- Ed. 128, 11437 (2016). based nanoparticles. Adv. Drug Deliv. Rev. 64, 1738–1748 (2012). 3. Ling, X., Wang, H., Huang, S., Xia, F. & Dresselhaus, M. S. The renaissance of 35. Lee, Y. K., Choi, E. J., Webster, T. J., Kim, S. H. & Khang, D. Effect of the black phosphorus. Proc. Natl Acad. Sci. USA 112, 4523 (2015). protein corona on nanoparticles for modulating cytotoxicity and 4. Gusmão, R., Sofer, Z. & Pumera, M. Black phosphorus rediscovered: from immunotoxicity. Int. J. Nanomed. 2015,97–113 (2014). bulk material to monolayers. Angew. Chem. Int. Ed. 56, 8052–8072 (2017). 36. Gao, N. et al. Steering carbon nanotubes to scavenger receptor recognition by 5. Chen, W. et al. Black phosphorus nanosheet-based drug delivery system for nanotube surface chemistry modification partially alleviates NFκB activation synergistic photodynamic/photothermal/chemotherapy of cancer. Adv. Mater. and reduces its immunotoxicity. ACS Nano 5, 4581 (2011). 29, 1603864 (2017). 37. Old, L. J. Tumor necrosis factor (TNF). Science 230, 630–632 (1985). 6. Wang, H. et al. Ultrathin black phosphorus nanosheets for efficient singlet 38. Bartneck, M. et al. Rapid uptake of gold nanorods by primary human blood oxygen generation. J. Am. Chem. Soc. 137, 11376 (2015). phagocytes and immunomodulatory effects of surface chemistry. ACS Nano 4, 3073–3086 (2010).

10 NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-04873-7 ARTICLE

39. Reddy, S. T. et al. Exploiting lymphatic transport and complement activation Acknowledgements in nanoparticle vaccines. Nat. Biotechnol. 25, 1159 (2007). This work was written in memory of Dr. Qing Huang for his great inspiration for this 40. Nel, A., Xia, T., Mädler, L. & Li, N. Toxic potential of materials at the research. Financial support was provided by the National Science Foundation of China nanolevel. Science 311, 622 (2006). (21622103, 21571098 and 21671099), the Natural Science Foundation of Jiangsu Pro- 41. Nel, A. Air pollution-related illness: effects of particles. Science 308, 804 vince (BK20160022) and Shenzhen Basic Research Program (JCYJ20170413150538897 (2005). and GJHS20170310093035206). 42. Ge, C. et al. Binding of blood proteins to carbon nanotubes reduces cytotoxicity. Proc. Natl Acad. Sci. USA 108, 16968 (2011). Author contributions 43. Kang, B. et al. Carbohydrate-based nanocarriers exhibiting specific cell targeting with minimum influence from the protein corona. Angew. Chem. J.M and Q.X. carried out the experimental works; J.M. and W.W. wrote the manuscript; Int. Ed. 54, 7436 (2015). J.Z. and W.W. guided the project. 44. Peng, Q. et al. Preformed albumin corona, a protective coating for nanoparticles based drug delivery system. Biomaterials 34, 8521–8530 (2013). Additional information 45. Göppert, T. M. & Müller, R. H. Polysorbate-stabilized solid lipid nanoparticles Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- as colloidal carriers for intravenous targeting of drugs to the brain: 018-04873-7. comparison of plasma protein adsorption patterns. J. Drug Target 13, 179–187 (2005). Competing interests: The authors declare no competing interests. 46. Dobrovolskaia, M. A. & Mcneil, S. E. Immunological properties of engineered nanomaterials. Nat. Nanotechnol. 2, 469–478 (2007). Reprints and permission information is available online at http://npg.nature.com/ 47. Schöttler, S., Landfester, K. & Mailänder, V. Controlling the stealth effect of reprintsandpermissions/ nanocarriers through understanding the protein corona. Angew. Chem. Int. Ed. 55, 8806 (2016). Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in 48. Monopoli, M. P. et al. Physical−chemical aspects of protein corona: relevance published maps and institutional affiliations. to in vitro and in vivo biological impacts of nanoparticles. J. Am. Chem. Soc. 133, 2525–2534 (2011). 49. Wiśniewski, J. R., Zougman, A., Nagaraj, N. & Mann, M. Universal sample preparation method for proteome analysis. Nat. Methods 6, 359–362 (2009). Open Access This article is licensed under a Creative Commons 50. Pi, E. et al. Mechanisms of soybean roots’ tolerances to salinity revealed by Attribution 4.0 International License, which permits use, sharing, proteomic and phosphoproteomic comparisons between two cultivars. Mol. adaptation, distribution and reproduction in any medium or format, as long as you give – Cell. Proteom. 15, 266 288 (2016). appropriate credit to the original author(s) and the source, provide a link to the Creative fi 51. Uhlmann, T. et al. A method for large-scale identi cation of protein arginine Commons license, and indicate if changes were made. The images or other third party – methylation. Mol. Cell. Proteom. 11, 1489 1499 (2012). material in this article are included in the article’s Creative Commons license, unless 52. Zhu, Y. et al. Protein corona of magnetic hydroxyapatite scaffold improves cell indicated otherwise in a credit line to the material. If material is not included in the proliferation via activation of mitogen-activated protein kinase signaling article’s Creative Commons license and your intended use is not permitted by statutory pathway. ACS Nano 11, 3690 (2017). regulation or exceeds the permitted use, you will need to obtain permission directly from 53. Götz, S. et al. High-throughput functional annotation and data mining with the copyright holder. To view a copy of this license, visit http://creativecommons.org/ the Blast2GO suite. Nucleic Acids Res. 36, 3420 (2008). licenses/by/4.0/. 54. Mo, J., He, L., Ma, B. & Chen, T. Tailoring particle size of mesoporous silica nanosystem to antagonize glioblastoma and overcome blood–brain barrier. ACS Appl. Mater. Interfaces 8, 6811–6825 (2016). © The Author(s) 2018 55. Vizcaíno, J. A. et al. 2016 update of the PRIDE database and its related tools. Nucleic Acids Res. 44, D447–D456 (2016).

NATURE COMMUNICATIONS | (2018) 9:2480 | DOI: 10.1038/s41467-018-04873-7 | www.nature.com/naturecommunications 11